Alternative Measures of Standard Deviation Coefficient of Variation and Standard Error

The Paper examines the performances of some popular and frequently used measures of dispersion such as standard deviation, coefficient of variation and standard error and survey that they may not perform as good as anticipating in the presence of non-normality or outliers. The usefulness of the proposed measures is scrutinized with the frequently used measures of dispersion by bootstrap and jackknife computer based techniques as well as Monte Carlo simulation approach. In this paper, I propose new alternative measures of dispersion, namely DM-standard deviation, coefficient of deviation and DM-standard error. These measures should be fairly robust. The results demonstrate that DM-standard deviation, coefficient of deviation and DM-standard error outperforms than the standard deviation, coefficient of variation and standard error in a broad range of frequently occurring situations.